AI Automation & Integration Engineer in Bath, Michigan at BS&A Software

JOB OVERVIEW

The AI Automation & Integration Engineer builds, configures, and operates the shared technical foundation that enables BS&A teams to develop and run AI solutions safely, consistently, and cost-effectively. Reporting to the VP of AI Transformation and partnering closely with the CTO, this role provides access to approved models and APIs, enterprise integrations, reusable workflows and frameworks, shared runtime services, evaluation tooling, monitoring, and cost controls.

This is an integration, automation, and enablement role, not a model-research, model-training, or GPU-infrastructure role. This role connects approved models to BS&A's systems and data, creates reusable technical patterns, and helps teams move AI solutions into reliable production without rebuilding the same integrations for every use case.

BS&A deliberately designs for model flexibility rather than dependence on a single provider. Our advantage comes from the governance, engineering standards, and shared foundation that allow us to select the best approved model for each workload while managing cost, quality, security, and risk. This role builds and continually improves that foundation. The focus is internal. AI in our products is owned elsewhere.

The engineer configures and integrates existing capabilities first and develops custom solutions when nothing available does the job. It develops enterprise AI solutions directly when centralized ownership is appropriate and enables departmental engineering teams to build on shared capabilities when local ownership is the better path. The role works closely with the AI Portfolio Analyst on production-solution reviews and with IT & Security on the data, infrastructure, and security requirements within which AI solutions must operate.

KEY RESPONSIBILITIES

Shared AI Capabilities & Reference Implementations

• Configure, integrate, and maintain BS&A's shared, model-agnostic AI capabilities, including access to approved models and provider APIs, MCP and other tool integrations, orchestration and workflows, reusable frameworks, shared runtime services, evaluation tooling, monitoring, and cost controls.
• Create reusable templates, patterns, components, and reference implementations that reduce duplication and accelerate delivery across departments.
• Build and maintain integrations with trusted company knowledge and data sources, including SharePoint, Confluence, Jira, Salesforce, and approved internal databases, preserving source permissions, access controls, auditability, and other applicable security requirements.
• Support model and provider upgrades, deprecations, and migrations while minimizing disruption to production solutions.

Solution Engineering & Delivery

• Configure what BS&A already owns or can buy, and build custom only when nothing available does the job.
• Develop enterprise AI solutions directly when centralized ownership is appropriate, and enable departmental engineering teams to build their own solutions on shared capabilities when it is not.
• Implement CoE-built solutions according to BS&A's AI engineering standards for architecture, model eligibility, evaluation, and integration.
• Support build, buy, and configure decisions through technical proofs of concept and analysis of cost, quality, latency, security, scalability, and maintainability.
• Give the AI Portfolio Analyst an early read on whether a candidate use case is feasible, and whether we already own something that does it.
• Select the least expensive approved model that reliably meets the workload's quality, latency, security, and maintainability requirements.
• Build automated evaluation, regression testing, and deployment checks into the AI solution-development lifecycle.

Production Readiness, Reliability & Monitoring

• Instrument production AI solutions to monitor quality, latency, usage, failures, reliability, and cost.
• Provide the technical readiness assessment for production reviews, complementing the business evidence compiled by the AI Portfolio Analyst and the data, infrastructure, and security assessments owned by IT & Security.
• Own application-level reliability and incident response for CoE-managed components, partnering with IT & Security on underlying infrastructure incidents.
• Prepare and present technical evidence and recommendations for production-release decisions with the CTO and IT & Security.
• Identify recurring production issues and convert them into reusable improvements, standards, or preventive controls.
• Produce the documentation, runbook, and access detail the owning team needs to support a solution after launch.
• Decommission retired solutions cleanly, disabling agents and connectors, revoking credentials, and removing stale data connections.

Standards Adherence & Governance Partnership

• Ensure CoE-built solutions conform to BS&A's AI engineering standards, and provide departmental engineering teams with the patterns and technical guidance needed to apply the same standards.
• Partner with IT & Security and applicable data owners to implement controls based on data classification, access requirements, and security risk.
• Design against prompt injection wherever retrieved content can influence agent behavior, constraining tool access and agent permissions accordingly.
• Provide technical assessments for the Sanctioned AI Tool List and approved-model register.
• Surface cost, reliability, security, and technical-debt risks through the CoE's governance and portfolio-review processes.

Enablement & Knowledge Sharing

• Create and maintain internal technical documentation, implementation patterns, reference architectures, and training materials for approved AI development tools and practices.
• Run office hours and provide hands-on technical support for departmental engineering teams building on shared capabilities.
• Act as a technical resource for AI Champions and departmental engineers who are scoping, evaluating, or building AI solutions.

QUALIFICATIONS

• Three or more years of professional software, platform, or integration engineering experience, including demonstrated ownership of production systems.
• Hands-on experience delivering AI-enabled applications or workflows using LLM APIs from Anthropic, OpenAI, or comparable providers.
• Experience with retrieval and knowledge integration, tool calling, AI orchestration, or other AI-application development patterns.
• Demonstrated ability to build integrations and APIs against enterprise systems such as SharePoint, Confluence, Jira, Salesforce, or comparable platforms.
• Working knowledge of Microsoft Azure and cloud infrastructure, CI/CD, application deployment, monitoring, and observability practices.
• Hands-on experience with at least one low-code automation or agent-orchestration environment such as Copilot Studio, Power Platform, n8n, or Zapier.
• Understanding of data classification, authentication, authorization, access control, secrets management, and secure integration patterns.
• Demonstrated ability to operate effectively in new and undefined spaces, create reusable patterns from ambiguity, and adapt as tools and technical requirements change.
• Strong collaboration and communication skills, with the ability to work with engineers, architects, business Champions, IT & Security partners, and non-technical stakeholders.
• Demonstrated judgment about where AI is appropriate, where deterministic software is the better solution, and how to balance experimentation with production reliability.
• Bachelor's degree in Computer Science, Software Engineering, or a related field, or equivalent demonstrable professional experience.

PREFERRED QUALIFICATIONS

• Experience implementing human-in-the-loop workflows, including approval steps that run before an AI solution acts.
• Experience choosing an existing managed or purchased capability over building something custom.
• Practical experience with Model Context Protocol or comparable tool-integration protocols and agent frameworks.
• Familiarity with Claude and the Anthropic ecosystem, including Claude Enterprise or Claude Code, or equivalent depth with another frontier-model provider.
• Experience building retrieval-augmented generation, enterprise search, or knowledge-grounded AI applications.
• Experience with cloud identity and access management, secrets management, infrastructure as code, or containerized application deployment.
• Experience in SaaS, govtech, fintech, the public sector, or another environment involving sensitive customer data.
• Experience in or adjacent to a Center of Excellence, platform-engineering, internal-developer-platform, or developer-enablement function.